IN THE COMPETITION COMMISSION OF SOUTH AFRICA In the complaint submitted by:
TREATMENT ACTION CAMPAIGN
Concerning the conduct of:
MSD (PTY) LTD
MERCK & CO., INC. AND RELATED COMPANIES
STATEMENT OF COMPLAINT
THE COMPLAINANT
1. The complainant is the TREATMENT ACTION CAMPAIGN (“the TAC”), an association
incorporated under section 21 of the Companies Act 61 of 1973 with its head office at 34
Main Road, Muizenberg, Cape Town.
2. As set out in its Constitution, the objectives of the TAC include (but are not limited to) –
a) Campaigning for equitable access to affordable treatment for all people with
HIV/AIDS;
b) Challenging by means of litigation, lobbying, advocacy and all forms of
legitimate social mobilisation, any barrier or obstacle that limits access to
treatment for HIV/AIDS in the private and public sector; and
c) Campaigning for access to affordable and quality health care for all.
3. Further detail on the TAC – including its work regarding access to medicines in general
and antiretroviral (“ARV”) medicines in particular – is set out in the affidavit of
Abdurrazack “Zackie” Achmat, the organization’s chairperson and a person living openly
with HIV. Achmat’s affidavit – attached marked Annexure TAC – also sets out the
TAC’s previous interactions with the Competition Commission (“the Commission”). In
2
particular, it details the TAC’s role in the 2002 complaint brought by Hazel Tau and
others against the GlaxoSmithKline (“GSK”) and Boehringer Ingelheim (“BI”) groups of
companies in case number 2002Sep226 (“the Tau case”).
THE SUBJECTS OF THE COMPLAINT
4. The first subject of the complaint is MSD (PTY) LTD (“MSD”), a company duly
incorporated under the laws of South Africa and with its registered office at 117 16th
Road, Halfway House. MSD is cited because to the best of the complainant’s
knowledge it has and exercises the exclusive right – subject to its licensing agreements
with Aspen Pharmacare Holdings Limited (“Aspen”) and Adcock Ingram Limited
(“Adcock”) – to market and sell the ARV medicine efavirenz (“EFV”) in South Africa
(branded as Stocrin®), as the South African representative of Merck & Co., Inc.
5. The complaint is also submitted against MERCK & CO., INC. (“Merck”), a research-
based pharmaceutical company with headquarters at One Merck Drive, Whitehouse
Station, NJ 08889-0100, USA, and/or those companies that are related to it and have
the right to market and/or sell EFV to any entity in South Africa (including MSD). As the
complainant has been unable to ascertain which particular companies related to Merck
have the right to sell and market EFV to or in South Africa, it requests that the
Commission obtain this information during the course of its investigations.
6. For convenience, MSD and Merck – as well as any companies that are related to Merck
and have the right to market and/or sell EFV to any entity in South Africa (including
MSD) – are referred to collectively as the respondents.
7. EFV is claimed per se in South African Patent No. 93/5724, which is currently in force
and will expire in the normal course on 6 August 2013. As at July 2007, Merck was
registered as the relevant patentee, with its address for service under section 87 of the
Patents Act 57 of 1978 being DM Kisch Inc., 54 Wierda Road West, Wierda Valley,
Sandton.
3
8. The invention of South African Patent No. 93/5724 also covers pharmaceutical
compositions containing EFV; its use in the treatment of HIV/AIDS; its use in
combination with the ARV medicines zidovudine (“AZT”), didanosine (“ddI”) and
zalcitabine (“ddC”); and a process for synthesizing EFV.
9. There is a further South African patent that covers combinations of EFV and other ARV
medicines. As at July 2007, Gilead Sciences Inc. (“Gilead”) – a researched-based
pharmaceutical company with headquarters at 333 Lakeside Drive, Foster City, CA
94404, USA – was registered as the patentee of South African Patent No. 2005/05852,
with its address for service under section 87 of the Patents Act 57 of 1978 being
Bowman Gilfillan Attorneys, 165 West Street, Sandton.
10. South African Patent No. 2005/05852, which is in force and will expire in the normal
course on 13 January 2024, covers the use of the combination of the ARV medicines
tenofovir disoproxil fumarate (“TDF”) and emtricitabine (“FTC”) as anti-HIV therapy. Of
relevance to this complaint is that the patent also claims the combination of TDF, FTC
and EFV, and an oral pharmaceutical dosage form comprising TDF, FTC and EFV.
11. However, Gilead has expressly committed itself not to enforce whatever HIV-related
patents it holds in the 95 developing countries (including South Africa) eligible for its “no-
profit pricing”. In addition, Merck has the exclusive right to market TDF/FTC/EFV in
South Africa. In terms of an agreement between Gilead and Merck, “Gilead will
manufacture ATRIPLA using efavirenz supplied by Merck … [which] in turn will handle
distribution of the product in the countries covered by the agreement”. This is set out in
a press release dated 11 August 2006, which is available online at
http://www.gilead.ca/wt/sec/pr_895234 and is attached marked Annexure MSD1.
THE ESSENCE OF THE COMPLAINT
12. The complainant alleges that the respondents have violated section 8(c) of the
Competition Act 89 of 1998 (“the Act”) by refusing to license –
4
a) Any existing company in South Africa (other than Aspen and Adcock) to import
into, manufacture, use, offer to dispose of and/or dispose of in South Africa,
generic EFV products; and
b) Any existing company in South Africa (including Aspen and Adcock) to import
into, manufacture, use, offer to dispose of and/or dispose of in South Africa, co-
formulated and/or co-packaged generic products containing EFV and at least
one other ARV medicine.
13. Although MSD claims that it is “currently assessing additional requests for a license to
manufacture and distribute efavirenz in South Africa” (Annexure JMB20, which is
attached to Jonathan Berger’s affidavit marked Annexure JMB), it has already refused
to license at least two existing companies in South Africa:
a) In a letter to Cipla-Medpro (Pty) Ltd (“Cipla-Medpro”) dated 9 May 2006, MSD’s
former Chief Executive Officer (“CEO”) explained his company’s decision not to
license Cipla-Medpro:
As I previously advised, Merck & Co considers each request for a license seriously.
We evaluate such requests based on a thorough assessment of the demand for
Efavirenz in South Africa in South Africa as well as other criteria which were
previously communicated to Cipla …. We are monitoring the demand for Stocrin both
in public and private sectors, and have concluded that at this stage MSD SA is in
position to adequately meet all of the demand for this product. In addition, we expect
Aspen, another supplier, will be in a position to come to market in a few months to
ensure multiple supplies for this medicine. (Annexure MSD2)
b) More recently, Cipla-Medpro was once again effectively denied a licence. In a
letter dated 25 September 2007, MSD’s new CEO sought to justify his
company’s position as follows:
Given that the granting of the second efavirenz licence occurred only recently at the
end of August 2007, further applications for efavirenz licences will be evaluated and
the due process will be followed once a clearer understanding of the supply of
efavirenz has been established. (Annexure MSD3, emphasis added)
5
c) The complainant has reliably been advised that MSD has also refused to license
Sonke Pharmaceuticals (Pty) Ltd, a joint venture between Ranbaxy (SA) (Pty)
Ltd and Community Investment Holdings (Pty) Ltd.
14. Regarding MSD’s effective refusal to license, a letter (dated 4 June 2007) sent by the
complainant’s legal representatives – the AIDS Law Project (“ALP”) –– to which MSD
has yet to reply, records as follows:
2. While we appreciate your timely response, we do not believe that you have addressed
the key issues raised in our letter dated 21 May 2007. In particular, you have not
committed MSD to –
a. Licensing additional generic companies to supply the South African and SADC
[Southern African Development Community] markets on reasonable terms; and
b. Permitting Aspen to bring co-formulated and/or co-packaged generic products
containing EFV and at least one other ARV medicine to market.
3. Instead, you have simply –
a. Asserted that you “are currently assessing additional requests for a license to
manufacture and distribute efavirenz in South Africa”, without committing to
issuing licences should the prospective licensees be able to satisfy your
concerns relating to certain objective criteria – “[i]ssues of bioequivalence,
safety, quality, forecasting and planning”; and
b. Postponed any decision relating to licences in respect of fixed-dose combination
(FDC) products containing EFV, further claiming that the MCC does not permit
co-packaging (more on this below).
The ALP letter is attached to Berger’s affidavit as Annexure JMB21.
15. As already indicated, MSD has subsequently granted a second licence – apparently on
substantially similar (if not the same) terms as the Aspen licence – to Adcock. The
award of that licence, which is similarly restricted to stand-alone EFV products, is
6
addressed in a 7 September 2007 press release attached marked Annexure MSD4 and
available online at http://www.adcock.com/article.aspx?ArticleId=43.
16. By refusing to license further existing companies in South Africa, and by refusing to
permit Aspen and Adcock to bring co-formulated and/or co-packaged generic products
containing EFV and at least one other ARV medicine to market, the respondents have –
without good cause – collectively threatened access to comprehensive treatment for
HIV/AIDS in both public and private sectors by –
a) preventing the market entry in South Africa of significantly cheaper generic EFV
products;
b) preventing the market entry in South Africa of a range of co-formulated and/or
co-packaged products containing EFV; and
c) placing the sustainability of supply of EFV products in South Africa under threat.
STRUCTURE OF THIS STATEMENT OF COMPLAINT
17. This statement of complaint addresses the following two factual issues:
a) The extent, nature and treatment of HIV infection in South Africa; and
b) The implications of the respondents’ refusal to license.
The legal arguments made in support of this statement of complaint are set out in a
separate document entitled LEGAL SUBMISSIONS.
18. The statement of complaint first addresses the following in relation to the extent, nature
and treatment of HIV infection in South Africa:
a) State of the HIV/AIDS epidemic in South Africa; and
b) Treating HIV infection with highly active ARV therapy (“HAART”).
7
19. The statement then addresses the implications of the respondents’ refusal to license, in
relation to –
a. the exclusion of significantly cheaper generic EFV products from the South
African market;
b. the exclusion of a range of co-formulated and/or co-packaged ARV products
from the South African market; and
c. threats to the sustainability of supply.
EXTENT, NATURE AND TREATMENT OF HIV INFECTION IN SOUTH AFRICA State of the HIV/AIDS epidemic in South Africa
20. Because of its extent, nature and impact, HIV/AIDS represents the “greatest threat to
public health in our country” and “the most important challenge facing South Africa since
the birth of our new democracy”. (Minister of Health v Treatment Action Campaign (No
2) 2002 (5) SA 721 (CC) at paragraphs 93 and 1 respectively). The impact of the
epidemic is discussed in frank terms in the report A Nation in the Making – A discussion
document on macro-social trends in South Africa, published by the Presidency in 2006
and available online at http://www.info.gov.za/otherdocs/2006/socioreport.pdf:
[T]here is clearly not only a pandemic in silent attack, but its fatal impact is starting to
express itself palpably in both morbidity and mortality. The most affected in this regard are
able-bodied citizens in the prime of their lives. These would most likely be parents of young
children and possibly breadwinners of extended families who are also among the most
skilled within the population (at page 65).
21. Despite this reality, government was for some time reluctant to act with the urgency and
commitment demanded by the epidemic. Finally, after many years of conflict with civil
society, strong political leadership from Deputy President Phumzile Mlambo-Ngcuka and
former Deputy Minister of Health Nozizwe Madlala-Routledge led to the adoption of the
new national HIV & AIDS and STI Strategic Plan for South Africa 2007-2011 (“the NSP”)
in 2007 – endorsed by the South African National AIDS Council (“SANAC”) on 30 April
8
2007 and thereafter approved by Cabinet on 2 May 2007. A copy of the executive
summary of the NSP is attached marked Annexure MSD5. The complete document will
be made available upon request.
22. All key stakeholders, including government’s most vociferous critics in civil society, were
integrally involved in the NSP’s conceptualisation, development and finalisation.
Importantly, the NSP affirms the necessity of dealing decisively, comprehensively and
urgently with the epidemic, noting as follows (at page 17):
HIV and AIDS is one of the major challenges facing South Africa today. Some two decades
since the introduction of this disease in the general population, the epidemiological situation
is characterized by very large numbers of people living with HIV and a disproportionate effect
on particular sectors of society, viz.: young women, the poor, as well as those living in
underdeveloped areas in the country. HIV and AIDS, however, affect the lives of all people
who live in South Africa in different ways.
23. The primary aims of the NSP are to reduce the rate of new HIV infections by 50% by
2011; and to reduce the impact of HIV/AIDS by expanding access to appropriate
treatment, care and support to 80% of those in need by 2011. The four priority areas of
the NSP are prevention; treatment, care and support; research, monitoring and
surveillance; and human rights and access to justice. They collectively identify a range
of activities, targets and programmes required to ensure that the NSP’s aims are
realised.
24. As an evidence-based plan, the NSP places much reliance on the report Demographic
Impact of HIV/AIDS in South Africa (“Demographic Impact”), the most comprehensive
analysis of the extent and impact of the South African epidemic yet undertaken. A copy
of the Executive Summary of Demographic Impact is attached marked Annexure MSD6. The full report, which will be made available upon request, is available online at
http://www.assa.org.za/aids/content.asp?id=1000000449.
25. Demographic Impact makes use of the ASSA2003 AIDS and Demographic model, which
in turn “is based on a thorough analysis of a range of epidemiological and demographic
9
data including the antenatal surveys and recorded deaths up to the year 2003. In
addition the projections allow for the impact of major current interventions.” (Page i)
26. In his expert affidavit (Annexure LJ), the University of Cape Town’s Leigh Johnson –
one of the authors of Demographic Impact – makes use of the ASSA2003 AIDS and
Demographic model to consider the impact of the HIV/AIDS epidemic and its
implications for the provision of ARV treatment in accordance with the targets and
timeframes set out in the NSP. On projections regarding mortality and life expectancy in
particular, Johnson comes to the following conclusions (at paragraphs 29 – 32):
As a result of the low rates of access to treatment and the high numbers of people sick with
AIDS, South Africa is experiencing high levels of AIDS mortality. It is estimated that during
2006, approximately 350,000 people died from AIDS. This has led to a reduction in life
expectancy at birth, from approximately 63 years in 1990, to approximately 51 years in 2006.
…
If the mortality rates experienced in 2006 were to continue in future, roughly 56% of current
15-year olds would die before reaching the age of 60. The proportion would be 29% if
mortality rates continued at levels experienced in 1990. …
AIDS has thus led to massive increases in mortality rates, particularly in the economically
active ages. …
As a result of the rapid increase in adult mortality levels, there has been a dramatic rise in
numbers of orphaned children. The number of children under the age of 18 who have lost
both parents is estimated at 430,000 as at mid-2006. A further 1.1 million have lost their
mother, and another 2.4 million have lost their father. Equivalently, 2.4% of children have lost
both parents, a further 6% have lost their mother, and another 13% have lost their father.
Treating HIV infection with HAART
27. HIV/AIDS can be treated effectively by ensuring access to ARV medicines, which
specifically target HIV infection itself rather than AIDS-related opportunistic infections.
These drugs are combined in various different treatment regimens that collectively are
known as HAART. In this statement of complaint, which draws heavily on Dr Robin
10
Wood’s expert affidavit (attached marked Annexure RW), HAART and ARV treatment
are used interchangeably.
28. HAART has revolutionised the management of HIV infection, resulting in a radical
reduction in mortality and morbidity figures amongst groups with access to treatment.
Additional health benefits include the reduction and/or elimination of opportunistic
infections, the restoration of immune function, and a reduction in infectiousness. With
access to HAART, people with HIV/AIDS are able to lead longer and healthier lives.
Access to ARV treatment thus directly results in an improved quality of life and the
restoration of dignity, allowing people with HIV/AIDS who were previously ill to resume
ordinary everyday activities, such as work.
29. Since 1996, HAART has been the standard of care in Europe, North America and Brazil.
Scientific consensus confirms that at least three drugs should be used together to
ensure maximum clinical efficacy, reduce side effects and limit the emergence of
resistant strains of the virus. The simultaneous use of three or more drugs, including
drugs from different therapeutic classes, is essential for individual clinical outcomes and
for public health protection. In short, it is necessary to have access to a combination of
drug choices both within and between drug classes.
30. ARV medicines target either a particular step in the life cycle of HIV or its interaction with
host cells. The ARV medicines in general use in South Africa inhibit one of two key viral
enzymes required by HIV for replication – reverse transcriptase (essential for the
completion of the early stages of viral replication) or protease (required for the assembly
and maturation of new HIV) – and can be divided into the following therapeutic classes:
a) Nucleoside analogue reverse transcriptase inhibitors (“NRTIs”): AZT, lamivudine
(“3TC”), FTC, abacavir (“ABC”), stavudine (“d4T”) and ddI;
b) Nucleotide analogue reverse transcriptase inhibitors (“NtRTIs”): TDF;
c) Non-nucleoside reverse transcriptase inhibitors (“NNRTIs”): nevirapine (“NVP”)
and EFV;
11
d) Protease inhibitors (“PIs”): lopinavir (“LPV”), ritonavir (“RTV”), indinavir (“IDV”),
saquinavir (“SQV”) and atazanavir (“ATZ”).
31. Some ARV medicines are available in fixed-dose combination form. LPV/r is the only
form in which LPV is manufactured internationally. In South Africa, the only
combinations currently available other than LPV/r are TDF/FTC, AZT/3TC, 3TC/ABC,
AZT/3TC/ABC and d4T/3TC/NVP. TDF/FTC is the only form in which FTC is marketed
in South Africa, although it is available as a stand-alone product outside of the country.
32. There is a range of other ARV medicines – such as fosamprenavir, tipranavir,
enfuvirtide, darunavir, raltegravir and maraviroc – that are available elsewhere but have
yet to be registered for use in South Africa. Certain combinations of ARV medicines –
such as AZT/3TC/NVP, TDF/FTC/EFV, TDF/3TC/EFV and d4T/3TC – are available for
use in countries other than South Africa. A few ARV medicines – such as amprenavir
and zalcitabine (“ddC”) – are no longer available for use anywhere. In addition, all stock
of one ARV medicine – nelfinavir (“NFV”) – was recently recalled as a result of product
contamination.
33. The ability to devise different drug regimens allows practitioners to take into account
problems such as proven antagonism between specific drugs, potency and side effect
profile, potential for maintenance of future treatment options, pregnancy, potential for
primary acquisition of resistant viral strains, the incidence of tuberculosis (“TB”) and
hepatitis B and/or C, amongst other factors. All these factors have an influence on what
drug regimen is selected.
34. In the public sector, however, only a limited number of these ARV medicines (AZT, 3TC,
ABC, d4T, ddI, NVP, EFV, LPV/r and RTV) are available for treating HIV infection. The
National ARV Treatment Guideline, to which Wood’s expert affidavit refers, details the
standard treatment regimens that govern the provision of HAART in the public sector.
These regimens are discussed in more detail in Wood’s affidavit.
12
35. The availability of a number of ARV medicines in a particular therapeutic class does not
necessarily mean that people with HIV/AIDS have a choice of products. The science of
HIV treatment shows that ARV medicines are generally not interchangeable. Numerous
factors combine to limit or completely exclude treatment combinations and options. The
implications of this are addressed in the legal submissions, most importantly in relation
to market definition and the establishment of dominance.
IMPLICATIONS OF THE RESPONDENTS’ REFUSAL TO LICENSE
36. Directly and through its legal representative, the complainant has engaged with the
respondents over many years regarding the need for them to license manufacturers
and/or importers of generic EFV medicines on reasonable and non-discriminatory terms.
Notwithstanding this engagement, the detail of which is set out in Berger’s affidavit, the
respondents’ conduct continues to give rise to the following three problematic outcomes:
a) The exclusion of significantly cheaper generic EFV products from the South
African market;
b) The exclusion of a range of co-formulated and/or co-packaged ARV products
from the South African market; and
c) Threats to the sustainability of supply.
These three matters are now addressed in turn.
Exclusion of cheaper generic EFV products from the market
37. Whilst the complainant does not allege that the respondents are engaged in prohibited
excessive pricing, it is nevertheless concerned about the impact – or potential impact –
of the respondents’ conduct on the pricing of EFV products respectively. Importantly,
the respondents claim that they are selling the relevant ARV medicines in South Africa
at cost. In other words, they claim to be making no profit on domestic sales of these
products. In particular, MSD claims that “STOCRIN [EFV] is already provided in South
13
Africa at prices from which MSD does not profit.” (Annexure JMB20 to Berger’s
affidavit)
38. The question, therefore, is not whether the respondents are able to bring the relevant
EFV products to market at lower prices, but rather whether the Act entitles them to
prevent generic companies from entering the market in these circumstances, namely
where such conduct:
a) will have little to no impact on the respondents’ profit margins;
b) excludes cheaper generic products from the market; and
c) results in threatening access to comprehensive treatment for HIV/AIDS in both
public and private sectors.
This issue is addressed in the legal submissions.
Pricing of EFV products
39. MSD markets stand-alone EFV products in South Africa at its best international prices,
which are offered to both the public and private sectors:
a) EFV 200mg – US$394.20 per adult per year; and
b) EFV 600mg – US$237.25 per adult per year (Annexure MSD7).
40. In addition, it has committed to pricing TDF/FTC/EFV at US$613.20 per adult per year in
developing countries such as South Africa (Annexure MSD7). To date, however, the
combination has yet to be registered here. It is therefore not generally available for sale
in the country
41. MSD’s prices – which appear to be determined by Merck – are significantly cheaper
today than they were at the beginning of the decade. Interestingly, Merck first
announced a substantial reduction in the price of EFV (200mg capsules) in March 2001
(Annexure MSD8), only days after the complainant had been admitted as amicus curiae
14
by the Transvaal High Court in the case of Pharmaceutical Manufacturers’ Association
and Others v Government of the Republic of South Africa and Others (case no:
4183/98). At the same time, Merck also reduced the price of IDV. Reuters reported that
“[t]he new prices represented reductions of some 40 to 50 percent from already
discounted prices pledged to African governments” in 2000 (Annexure MSD9). If this is
correct, EFV had previously been sold for somewhere between US$830 and US$1,000
for a year’s supply for one adult.
42. Since then, Merck has announced the following price reductions of EFV:
a) 23 October 2002: US$346.75 (600mg)
b) 7 March 2006: US$394.20 (200mg) and US$277.40 (600mg)
c) 14 February 2007: US$237.25 (600mg).
These prices are set out in Annexure MSD10, Annexure MSD11 and Annexure MSD12 respectively.
43. Reputable generic companies have ordinarily been able to undercut Merck’s prices.
Most of the following price data has been sourced from various editions of the Médecins
Sans Frontières (“MSF”) document entitled Untangling the web of price reductions: a
pricing guide for the purchase of ARVs for developing countries (“Untangling the web”).
Only the relevant pages of the various editions have been attached. The various
editions of the document, which are available online at http://www.accessmed-
msf.org/prod/view.asp?catid=1&, will be made available upon request.
a. In September 2001, when Merck’s price was US$500, the best generic price
was US$ 485 (Annexure MSD13).
b. In December 2002, when Merck’s price for 200mg EFV was still US$500, the
best generic price had dropped to US$438 (Annexure MSD14). At that time,
Merck was offering the lowest price for 600mg EFV.
15
c. By July 2006, when Merck’s prices for EFV were at US$394.20 (200mg) and
US$277.40 (600mg) respectively, the best generic prices had dropped to
US$225 and US$217 respectively (Annexure MSD15).
d. Merck’s February 2007 price reduction of 600mg EFV to US$237.25 was
subsequently bettered by generic companies, which have agreed to supply the
Clinton Foundation HIV/AIDS Initiative (“CHAI”) Procurement Consortium with
600mg EFV at US$164 – a saving of almost 31% when measured against
Merck’s current best price. The relevant CHAI document is attached marked
Annexure MSD16.
e. The lowest price for generic 200mg EFV now stands at US$185 – less than half
the current price of Merck’s equivalent product (Annexure MSD17).
44. Generic companies are also able to undercut Merck’s US$613.20 price for the
combination TDF/FTC/EFV (Annexure MSD16):
a) In terms of a recent agreement with the CHAI Procurement Consortium, Matrix
Laboratories has agreed to supply a generic version of the combination for
US$385 per adult per year – a saving of over 37% when measured against
Merck’s price.
b) Matrix also produces a similar combination that replaces FTC with 3TC.
According to Wood, FTC and 3TC are the only two ARV medicines that can
indeed be considered as substitutable for each other. The combination
TDF/3TC/EFV is available to the CHAI Procurement Consortium for US$339 per
adult per year. This represents a saving of almost 45% when measured against
Merck’s price for TDF/FTC/EFV.
45. Generic companies have been able to reduce their prices without having access to
significant economies of scale. Because of the price differential between EFV (US$164
for the best CHAI price) and NVP ($US45 for the best CHAI price), the NNRTI of choice
in most of the developing world is NVP. Brazil – which runs the largest public sector
ARV treatment programme outside of South Africa – has until very recently provided
16
access to EFV only in the form of Merck’s patented product. South Africa and Botswana
continue to do so.
46. In South Africa, more than two-thirds of all public sector patients accessing ARV
treatment are understood to be using EFV. This amounts to more than 230,000 people
currently using EFV as part of their ARV treatment in the public sector. Applying a
similar breakdown to the private sector yields an estimate of over 70,000 people on ARV
treatment who are currently taking EFV. These estimates of numbers of people on ARV
treatment in South Africa are based on the work of the Joint Civil Society Monitoring
Forum (“JCSMF”), described in Fatima Hassan’s affidavit which is attached marked
Annexure FH.
47. Implicit in Johnson’s expert affidavit is that we can estimate between 700,000 and
930,000 people to be using EFV as part of ARV treatment in the public sector by the end
of 2011. If generic companies had access to this already substantial and increasingly
expanding market, the marginal costs of production could drop significantly. With
sufficient competition, this would likely result in further price reductions.
48. Aspen, the first of MSD’s two licensees in respect of stand-alone EFV products, does
not appear to be one of the generic companies that are able and willing to undercut
Merck’s best prices. To date, the best price to which it has committed itself in the event
that it ever brings EFV products to market is US$240 for the 600mg tablet. It has made
no commitments whatsoever in respect of other EFV products (Annexure MSD18).
49. Adcock, MSD’s second licensee, has not been able to commit to any particular price.
Instead, its managing director has been reported as saying that Adcock will bring EFV to
market “at a significant discount to MSD’s version” (Annexure MSD19, available online
at http://www.adcock.co.za/Media_News.aspx?Year=2007&Month=7). Importantly, he
“was unable to provide an exact figure, saying it would depend to some extent on the
volumes sold.” The complainant is of the view that this constitutes more a wish than any
realistic or definite commitment.
17
50. If and when the two licensees bring their 600mg EFV tablets to the South African
market, they will only have to price their products marginally lower than MSD’s
US$237.25 price to capture significant shares of the private market. In the absence of
other licensees, they would not be under competitive pressure to lower their prices
further. In respect of the public sector tender, they may even be able to charge more
than MSD in the likely event that local manufacturers are afforded some margin in order
to further government’s stated industrial policy agenda. Only further licensees will be
able to ensure that the two licensees, both of whom produce locally, are compelled to
compete on the basis of price.
The need for cheaper ARV prices
The NSP on price reductions
51. The NSP identifies the need for steps to be taken to reduce the cost of ARV medicines
so as to ensure that it is indeed feasible to implement the plan. The low-cost scenario
(which sees only 60% of new AIDS cases accessing treatment by 2011) identifies ARV
treatment as responsible for 46% – adults (40%) and children (6%) – of the total cost of
the NSP (table 3 at page 117 of the NSP). The high-cost scenario (which sees 80% of
new AIDS cases accessing treatment by 2011, in line with one of the two primary goals
of the NSP) identifies ARV treatment as responsible for 51% – adults (44%) and children
(7%) – of the total cost of the NSP (table 4 at page 118 of the NSP).
52. In respect of price reductions, the NSP states as follows (at pages 116 and 121
respectively):
The cost implications of the NSP are large, in some options exceeding 20% of the health budget without considering the costs arising from the effect of the epidemic on hospital and primary care services. In attempting to increase the feasibility of this plan … [a]ttention should be placed on increasing the affordability of medicines.
…
18
It is estimated that, at current prices, the provision of antiretroviral therapy will account for
about 40% of the total cost of the NSP. This much needed service will soon be unaffordable
at the current drug prices.
53. The message in the NSP is clear: its ambitious targets can not be reached at the
current ARV medicine prices. An analysis of the first (and to date only) national ARV
medicine tender (Annexure MSD20) shows that the most appropriate and feasible way
to ensure that overall ARV medicine costs are reduced is by ensuring that the public
sector is able to procure cheaper EFV and/or LPV/r products. These medicines
collectively accounted for ± 63.9% of the value of the tender, even though only ± 24.3%
when measured in terms of volume. Collectively, the balance of ± 36.1% (in terms of
value) – accounting for ± 75.7% in terms of volume – was made up by seven other ARV
medicines. Unfortunately, generic companies internationally have not yet been able to
undercut the best international price charged for patented LPV/r, meaning that EFV
remains the only ARV medicine on the current tender list in respect of which significant
price reductions are at present possible.
54. Implicit in the NSP is that medically necessary and public health appropriate revisions to
public sector ARV treatment protocols may not be feasible if they would bring with them
significant cost implications. Put simply, cost may be the only factor preventing more
appropriate first-line regimens (such as TDF + FTC (or 3TC) + EFV) from being
introduced. At medicine costs of US$613.20 per adult per year (MSD’s TDF/FTC/EFV
product), the public sector may find it impossible to provide universal access to ARV
treatment. But at US$339 (Matrix Laboratories’ TDF/3TC/EFV), which is lower than the
current price paid by the public sector for regimen 1a (d4T + 3TC + EFV), a change is
indeed possible.
Limited access in the Western Cape
55. While there are numerous factors that currently limit access to ARV treatment in the
South African public sector, the complainant has been advised that but for the high cost
of EFV, many more patients in the Western Cape – which has residual capacity to
provide ARV treatment – would be able to benefit from this potentially life-saving medical
19
intervention. If the respondents were to license multiple generic companies to import
and/or manufacture EFV products, the public sector in the Western Cape would be able
to treat many more patients than it can currently do. Without access to cheaper ARV
medicines, the Western Cape Department of Health is unable to save the lives of many
patients with HIV/AIDS who are in need of treatment and cannot afford to access it in
the private sector.
56. Unfortunately, the complainant has been unable to access the official Western Cape
document in which this allegation is confirmed. It has been obliged to rely upon private
communications from concerned individuals working within or advising the provincial
government. The complainant therefore requests the Commission to approach the
Western Cape Department of Health for assistance in obtaining the relevant document.
If the Commission is unable to obtain it, the complainant would be willing to advise it
regarding whom in particular to contact.
Exclusion of co-formulated and/or co-packaged ARV products from the market
57. In his expert affidavit, Wood explains that the higher the pill burden – the number of
tablets, capsules or other dosage forms that a patient takes on a regular basis – the less
likely a person is to adhere to ARV treatment. Conversely, increased adherence is
associated with lower pill burdens. Wood further explains the necessity for high levels of
adherence to ARV treatment, and the negative impact of low levels of adherence on
individual treatment outcomes.
58. There are numerous ways in which the pill burden may be reduced. These include:
a) reducing the frequency of taking pills by –
i. using products with long-acting active pharmaceutical ingredients
(“APIs”); or
ii. developing sustained or extended release versions;
b) combining standard strengths of more than one API in a single pill, known as a
fixed-dose combination (“FDC”); or
20
c) co-packaging two or more separate pills in a single blister pack.
59. As already mentioned, two FDCs that contain EFV – TDF/FTC/EFV (to be marketed in
South Africa as AtriplaTM) and TDF/3TC/EFV – have already been developed. There
appears to be no scientific reason why EFV could not be combined in a single pill with
other ARV medicines that are also taken once daily and ordinarily form part of treatment
regimens containing EFV. If this were done, FDCs such as FTC/EFV, 3TC/EFV,
ddI/FTC/EFV and ddI/3TC/EFV could theoretically reach the market. But as indicated
above, the respondents’ refusal to license any generic company to import and/or
manufacture FDCs containing EFV and at least one other ARV medicine is an absolute
bar to products such as TDF/3TC/EFV reaching the South African market.
60. Some multinational generic companies – including those that have subsidiaries in South
Africa – market the following co-packaged products containing EFV in other countries:
a) AZT/3TC + EFV;
b) d4T/3TC + EFV; and
c) 3TC + ddI + EFV.
61. Once again, the respondents’ refusal to license on reasonable and non-discriminatory
terms constitutes an absolute bar to such products reaching the South African market.
Threats to the sustainability of supply
62. The Operational Plan for Comprehensive HIV and AIDS Care, Management and
Treatment for South Africa (“the Operational Plan”), adopted by Cabinet on 19
November 2003 (more recently incorporated into the NSP) and in terms of which ARV
treatment is provided in the public sector, recognises that a “central component of HIV
and AIDS care and treatment is the production, procurement and supply of medicines, in
particular antiretrovirals” (at page 143). The executive summary of the Operational Plan
is attached marked Annexure MSD21. The full plan is available online at
http://www.info.gov.za/issues/hiv/careplan.htm and will be made available upon request.
21
63. The Operational Plan recognises that in order to support its proper implementation, the
drug procurement system must – amongst other things – ensure that the supply of
medicines is “secure and sustainable at a volume large enough to meet the significant
demand envisioned” (at page 143). In particular, it recognises – at pages 145 and 146 –
that a multiplicity of suppliers is necessary to ensure sustainability of supply:
A number of viable and competing manufacturers will also guarantee security of supply
should any supplier fail for any reason. … There is always the risk of failure in the supply
chain of pharmaceuticals. It is intended that the procurement plan coordinates a sustainable
supply through the participation of viable suppliers, and local production of finished products
and active pharmaceutical ingredients.
64. In his expert affidavit, Wood explains what would happen to a person using EFV as part
of ARV treatment if the medicine in question were – for whatever reason – to become
unavailable. In his focus on ARV treatment that is ordinarily accessed in the public
sector, Wood comes to the following conclusions (at paragraphs 59 and 60):
For these reasons, in my considered expert opinion, if EFV were not available to health care
workers for the treatment of patients, it would have serious public health consequences. …
I should also point out that unintended consequences of interrupted treatment, irrespective of
the regimen are:
a) An undermining of adherence messages and patient confidence in the health care
system (why should patients be adherent when the health system is unable to
provide a sustainable supply of medicines?); and
b) An increased risk of AIDS and death, especially in those patients who had very low
CD4 counts at treatment initiation, where treatment interruption is for a significant
duration of time.
65. Johnson’s expert affidavit places Wood’s conclusions in context. He estimates that
between 1.25 and 1.54 million people will be on ARV treatment in the public sector by
the end of 2011, of which 204,000 and 149,000 respectively will be on second-line
treatment. In other words, between 1.046 and 1.391 million people will be on a first-line
22
regimen at that time. If the current pattern remains, with more than two-thirds of patients
in the public sector accessing EFV instead of NVP, we can estimate between 700,000
and 930,000 public sector patients using EFV.
Evidence of ex-manufacturer shortages
66. There are significant individual clinical and broader public heath implications if any ARV
medicine used in the public sector is not available (whether in branded or generic form).
There is a need for proactive steps to be taken to guard against any potential supply
problems in the future. In considering the need for such action, it is important to
appreciate that supply problems have already been experienced – in South Africa, in the
region, and internationally.
Shortages of EFV
67. Despite the respondents’ claims that sustainability of EFV supply is not a cause for
concern, there have been a number of occasions when there have been ex-
manufacturer shortages of the essential ARV medicine in the region. The complainant is
aware of shortages in at least two SADC countries: South Africa and Botswana.
68. In South Africa, shortages have been observed on at least three occasions in the public
sector. These shortages are discussed in Berger’s affidavit. Additional information
regarding EFV shortages is contained in Fatima Hassan’s affidavit. As is indicated in
Hassan’s affidavit, however, it may be necessary for the Commission – by using its
statutory powers of investigation – to obtain detailed information from the medical
schemes and disease management programmes (DMPs) identified regarding the
allegations of EFV shortages. For reasons not known to the complainant, the schemes
and DMPs appear hesitant to provide it with detailed information.
69. While the complainant is aware that there have been shortages of EFV in Botswana’s
public sector caused by Merck’s inability to supply product, it is unable to put such
evidence before the Commission. It will be able to provide the Commission with the
23
names and contact details of persons in Botswana who are well placed to provide such
evidence and to explain the impact of these EFV stockouts on the Botswana public
sector ARV treatment programme.
Shortages of other ARV medicines
70. The complainant has not attempted to conduct an exhaustive search of all ARV
medicine shortages that have been experienced in South Africa, the region or
internationally. Instead, it has focused largely on documented shortages in respect of
the ARV medicines that are the subject of this complaint. However, it is important to
note that the threat of a supply shortage is ever present. For this reason, two shortages
of other ARV medicines – both of which occurred in Botswana in late 2005 – are
considered below.
71. In a letter to the Deputy Permanent secretary in the Ministry of Health in Botswana
(dated 8 September 2005 and attached marked Annexure MSD22), Conrad Louw –
General Manager for GSK Southern Africa – described the reason behind his company’s
inability to deliver ARV medicines containing AZT timeously:
[D]ue to the augmented availability of funding from the International donor community over
the past couple of months, the global demand for ARV treatment has increased substantially
and unexpectedly, putting immense pressure on GlaxoSmithKline’s ability to meet new
orders within the existing supply dates. As a result of this unforecasted rise in demand, GSK
has been forced to extend the lead times for all new orders of ARVs containing zidovudine
from the current official level of 16 to 20 weeks.
72. Fortunately, the Government of Botswana was able to source alternative products
containing AZT from Aspen in South Africa. In settling the Tau complaint in 2002, GSK
had agreed to amend an existing licence it had already granted to Aspen so as to permit
exports of generic AZT and 3TC products to all sub-Saharan African countries (including
Botswana). Prior to the Tau case, Aspen’s licence from GSK only permitted it to sell to
the South African government.
24
73. At about the same time and as a result of an alleged “moderate shortage in the
availability of bulk stavudine globally” in or around October 2005, Bristol Myers-Squibb
(“BMS”) informed “customers” – including the Government of Botswana – that it
“expect[ed] to have limited stock [of d4T] available until mid 2006”. Disturbingly, BMS
suggested that new patients not be initiated on ARV treatment until the shortage had
been resolved. The relevant letter from BMS is attached marked Annexure MSD23.
74. GSK’s shortage of AZT products could be addressed because other suppliers had
indeed been licensed. The only way the d4T shortage could be addressed was by
limiting the numbers of patients accessing ARV treatment. Whilst additional licensees
would not have mattered in this case (given that the shortage was allegedly as a result
of the global shortage of bulk d4T), it does point to the need for additional licensees so
as to ensure access to alternative suppliers in the event of shortages as a result of a
particular manufacturer’s inability to supply.
Evidence of other supply problems
75. Thus far, this statement of complaint has described product shortages. A greater
concern is when a particular product is pulled from the market altogether. This may
happen when a product is recalled or when production ceases as a result of
manufacturing difficulties. In either case, the availability of generic alternatives is the
most appropriate means of averting harm to patients.
Product recalls
76. In a press release dated 14 June 2007 and entitled “WHO STATEMENT ON ROCHE'S
VIRACEPT® RECALL”, the World Health Organization (WHO) stated as follows:
“ROCHE informed WHO on 8 June 2007 of its global recall of its nelfinavir products
(Viracept®) …. The reason for the recall is identification of an impurity in some batches of
Viracept® …. The contaminant is a known genotoxic substance.”
25
According to Wikipedia, a genotoxic substance is a type of carcinogen, in particular one
that is “capable of causing genetic mutation and of contributing to the development of
tumors”. The WHO statement is attached marked Annexure MSD24.
77. In its statement, the WHO advised as follows:
Countries that have included Viracept® in post-exposure prophylaxis packs, should remove
and replace it with a suitable boosted protease inhibitor. If no boosted PI are available, dual
nucleoside therapy without a PI will remain effective.
Adults or children currently taking Viracept® should not interrupt their antiretroviral therapy. However, they should see their antiretroviral therapy provider as soon as possible, to change from Viracept® to a suitable alternative.
78. Due to the non-availability of generic NFV, patients using Viracept® had no option but to
change their treatment regimens. According to a report entitled “Near-global nelfinavir
(Viracept) recall 'very difficult situation'” (attached marked Annexure MSD25 and
available online at http://www.aidsmap.org.uk/en/news/EFA4601B-2827-4131-B2E8-
6EB722BEE1D9.asp), the recall had the following impact in four developing countries:
According to the Latin American News Agency Prensa Latina, Brazil’s Health Ministry –
which supplies nelfinavir at no cost to 9,000 adults and 250 children – told clinics to
immediately replace Viracept with Kaletra.
In Botswana, the Ministry of Health is to negotiate with Roche for a refund according to Dr
Loeto Mazhani, Director of Health Services in an interview with the Botswana Press Agency.
He said that although more than 85,000 individuals are on first-line antiretroviral therapy,
fewer than 1,000 patients were taking nelfinavir. He added that patients treated in
government or government assisted facilities have also been switched to Kaletra.
In Zambia, where Pelekelo Mwangisha, pharmaceutical evaluator of the Zambian
Pharmaceutical Regulatory Authority (PRA) told the Zambian Daily Mail that nelfinavir was
being taken by 100 patients, and that each patient’s doctor would have to decide which drug
to switch to. He also said that the PRA is undertaking spot checks in private health centres to
ensure that Viracept has been taken off the market.
26
Virawan Taengkaew, deputy secretary general of the Thai Food and Drug Administration
(FDA), told The Nation newspaper that the recall would affect very few people in Thailand. In
2006, 22 hospitals had prescribed nelfinavir to 190 individuals, but this year just 144 people
were taking the drug. She added that individuals taking nelfinavir could ask for compensation
if it is proven that they have been harmed by the contamination.
79. Because of the limited numbers of people using NFV as part of ARV treatment – not
more than 45,000 globally, with the vast majority living in the developed world – shifts to
other ARV medicines were indeed possible. This would not necessarily be so in the
case of a product recall, for example, of EFV from South Africa’s public sector. As can
be gleaned from Johnson’s affidavit, South Africa can expect to see at least 700,000
public sector users on EFV by the end of 2011. The public interest is thus best served
by ensuring that all reasonable steps are taken to guard against a product recall that
could result in such large patient numbers having to change their treatment regimens.
80. In contrast to the impact of the Viracept® recall, a product recall of Cipla-Medpro’s
AZT/3TC (“Duovir”) in South Africa in August 2004 – following concerns relating to the
quality of the bioequivalence studies that formed the basis of the medicine's registration
– did not result in the need for patients to change regimens. As a press release issued
by the complainant at the time noted:
People taking Duovir should know that there are other alternatives to this drug available in
South Africa. In addition to Duovir, three other versions of the AZT/lamivudine combination
have been registered by the MCC …. In addition, the MCC has also registered four versions
of each of the individual antiretroviral drugs AZT and lamivudine ….
Attached marked Annexure MSD26, the press release is available online at
http://www.tac.org.za/newsletter/2004/ns11_08_2004.htm.
Manufacturing problems
81. In July 1998, US-based pharmaceutical company Abbott Laboratories (“Abbott”)
experienced “manufacturing difficulties” with soft-gel capsule (“SGC”) RTV. In a letter to
27
health care providers in the United States, which is available online at
http://www.fda.gov/medwatch/safety/1998/norvir.htm and is attached marked Annexure MSD27, Abbott explained:
Abbott Laboratories is experiencing manufacturing difficulties with the capsule formulation of
our HIV protease inhibitor, Norvir (ritonavir), which will result in a shortage of capsules. We
have encountered an undesired formation of a Norvir crystalline structure that affects how
the capsule form of Norvir dissolves. It is our plan to supply Norvir oral solution (liquid) to
provide continued Norvir therapy for patients. Norvir capsules currently in distribution are not
affected by this issue. When used in accordance with the prescribing information, product on
the market is safe and effective.
82. Abbott’s original optimism regarding shortages appears to have been shortlived. As is
noted in “Ritonavir Users Put on Liquid Diet” (GMHC Treatment Issues, (July/August
1998) available online at http://www.thebody.com/content/art13463.html and attached
marked Annexure MSD28),”production problems had forced the company to
discontinue manufacture of the capsule formulation”:
Although the production problem first occurred in June, Abbott waited until the end of July to
inform the community, probably hoping to resolve the situation in time to avoid a shortage.
Unfortunately, by the time Abbott made its announcement, there was less than a month's
worth of capsules left. In order to prevent a run on the supply and stretch out the remaining
stock, Abbott controlled inventory by shipping wholesalers only their usual size orders. No
additional quantities were sent out nor special orders filled.
On the retail level, pharmacies scrambled to manage supply and demand. At least one major
drug store in New York City reported limited quantities of both capsules and liquid by the first
week of August and was rationing out seven-day allotments of the drug. Abbott is increasing
production of the oral formulation so that by the time pharmacies completely run out of the
capsules, there should be enough liquid to meet the demand.
83. SGC RTV did eventually return to the market, some months later. Despite its bad taste,
RTV liquid could be used as an alternative in the interim. But if the liquid had not been
available, many patients’ ARV treatment may have been disrupted completely. At that
28
stage, LPV/r had not yet reached the market. All other PIs – with the sole exception of
NFV – had to be taken with a boosting dose of RTV.
CONCLUSION
84. The TAC and ALP have engaged in a sustained campaign for many years to ensure that
MSD and Merck grant multiple non-exclusive voluntary licenses on reasonable and non-
discriminatory terms. In response, the respondents have granted two licences on
unreasonable grounds. Their failure to act appropriately is to be regretted, but can be
addressed, even at this late stage, to avert any future damage.
85. Should the Commission need any further information, it should not hesitate to contact
any of the individual deponents or the ALP, which has filed this complaint on behalf of
the complainant. In addition, the ALP will accept service of all documents on behalf of
the complainant. The ALP’s offices are located on the sixth floor of the Braamfontein
Centre, 23 Jorissen Street, Braamfontein, Johannesburg. It can be contacted at (011)
356-4100 (t) and (011) 339-4311 (f).
Top Related